The Effects of Reward Quality on Risk-sensitive Foraging Craft*, B.B., Church, A.C., Rohrbach, C.M.,

1
The Effects of Reward Quality on Risk-sensitive
Foraging Craft, B.B., Church, A.C., Rohrbach,
C.M., Bennett, J.M.
Introduction
Method
Results

• Subjects
• Male, Sprague Dawley (a laboratory strain of
  Rattus norvegicus) rats n20
• Apparatus
• The experimental procedure was conducted in
  Lafayette operant conditioning chambers housed in
  a sound attenuating box.
• The chambers (27.94 x 21.59 x 21.59 cm) had
  stainless steel grid flooring, stainless steel
  front and back walls, and a Plexiglas door,
  ceiling, and side walls.
• Two bar presses, equidistant to the feeding tray,
  corresponded with either a variable (risk-prone)
  or constant (risk-averse) reward option.
• Experimental Session
• The Sugar Group (SG, n10) chose between a
  risk-prone (1 or 5 pellets, p0.5) and a
  risk-averse (3 pellets, p1.0) option that
  delivered a calorically rich reward (45 mg
  BioServe sugar pellets)
• The Grain Group (GG, n10) chose between a
  risk-prone (1 or 5 pellets, p0.5) and
  risk-averse (3 pellets, p1.0) option that
  delivered a calorically poor reward (45 mg
  BioServe grain pellets).
• Delay to reward was held constant at .1s.
• Subjects were deprived to 90 of their
  free-feeding body weight.
• Once subjects reached the target weight, choice
  preference was measured using operant
  conditioning chambers.

Figure 1. The mean number of choices for the
variable option during the final three sessions
of the experiment.

• Risk-sensitive Foraging Theory
• Risk-sensitive Foraging Theory (RSFT) was
  developed to explain a foragers shift in choice
  between a variable (risk-prone) or constant
  (risk-averse) option. In typical RSFT studies, a
  risk-averse choice yields a constant return,
  whereas a risk-prone choice yields a variable
  return. If an organism displays a risk-prone or
  risk-averse choice bias, the organism is said to
  be risk-sensitive.
• Daily Energy Budget Rule
• The Daily Energy Budget (DEB) rule was developed
  to describe qualitative shifts in
  risk-sensitivity and makes the following
  assumptions first, there is a non-linear
  relationship between a foragers fitness and the
  rate of gain, and second, risk-sensitivity is
  contingent on the relationship between a
  metabolic reference and the rate of caloric
  return from a food source. For example, as a
  forager reaches a negative energy budget, the DEB
  rule describes the organism as seeking a reward
  with the greatest amount of return despite the
  variability of the reward.
• Past studies, primarily with insects, have
  manipulated reward quality by changing the mean
  reward amount or the concentration of calories in
  rewards, but support for the DEB rule in these
  experiments differs across species and procedures
  used. Contrary to insect studies, past
  experiments using rats have observed shifts in
  risk-sensitivity that are consistent with the DEB
  rule when the mean reward amount and response
  effort was manipulated.
• Given that rats displayed shifts in choice as the
  result of fluctuations in the mean reward amount,
  it is possible that changes in the rate of
  caloric return could result in differences in
  risk-sensitivity that are consistent with the DEB
  rule. However, no studies have been conducted
  using mammalian species where the mean reward
  amount and response effort was controlled and
  caloric return was manipulated.

Conclusion

• The results of the present study confirmed that
  subjects risk-sensitive choices were influenced
  by reward quality. In addition, these results are
  interesting in that they indicate that
  differences in the rate of caloric gain from a
  particular food source resulted in qualitative
  differences in risk-sensitivity in rats when the
  mean reward amount, delay to reward, response
  effort, and body mass were held constant.
• Choice preferences in the current experiment are
  consistent with the predictions of the DEB rule.
  The DEB rule describes qualitative changes in
  risk-sensitivity as being the result of a
  foragers fitness and the rate of gain from a
  food source. In the current experiment, subjects
  in the SG received rewards with a higher rate of
  caloric gain and displayed a risk-averse
  strategy. However, subjects that received rewards
  with a lower rate of calorie gain made more
  variable option choices than the GG and were
  risk-indifferent.
• One potential limitation of the current study was
  in regard to the type of rewards used. For
  instance, although the rate of caloric gain was
  different for the sugar and grain rewards,
  subjects in the current experiment could have
  been responding to the sweetness of the sugar
  pellets instead of the caloric value. Future
  studies should examine the potential effects of
  taste on risk-sensitivity by employing a sweet,
  nonnutritive option (i.e. saccharin or sucralose)
  along with a sugar and grain option.

Results

• All data were analyzed using an alpha level of
  .05.
• An Independent Samples t-Test was conducted to
  determine the difference between the frequency of
  risk-prone choices made by the SG (M4.76, SD4)
  and GG (M11.83, SD4.88) groups over the last
  five days of data collection.
• The Independent Samples t-Test revealed a
  statistically significant difference between the
  groups risk-prone choices, t(18)3.54, p.002, r
  2 .41 see Figure 1.
• Furthermore, a One-Sample t-Test was used to
  determine if the groups risk-prone responses
  deviate from chance alone. The test revealed a
  statistically significant choice preference for
  the SG, t(9) 4.14, p .003, r 2 .66. As for the
  GG, no statistically significant difference was
  observed, t(9) 1.19, p.27, r 2 .14.

Purpose

• We hypothesized that  subjects in receiving grain
  pellets would become more risk-prone due to the
  low caloric gain of the grain pellets, whereas
  subjects in the receiving sugar pellets would
  become more risk-averse due to the high caloric
  return of the sugar pellets.
• Furthermore, we hypothesized that subjects in
  both groups would display a risk-sensitive choice
  bias.  

Contact information craftb_at_spu.edu